Q Fever outbreak among staff at a South African veterinary hospital - epidemiological implications for One Health in Africa

¹Department of Companion Animal Clinical Studies, Faculty of Veterinary Science, University of Pretoria, Pretoria, South Africa, ²Department of Production Animal Studies, Faculty of Veterinary Science, University of Pretoria, Pretoria, South Africa, ³Thrive Family Practice Medical Practice, Ballito, South Africa, ⁴Department of Veterinary Tropical Diseases, Faculty of Veterinary Science, University of Pretoria, Pretoria, South Africa

^&Corresponding author
Karen Helena Keddy, Department of Veterinary Tropical Diseases, Faculty of Veterinary Science, University of Pretoria, Pretoria, South Africa

Introduction    

Coxiella burnetii, the intracellular bacterium causing Q-fever, is rarely reported in human populations in Africa. Limited studies have suggested a high prevalence in exposed populations, such as agricultural and abattoir workers. Amongst domestic animals, ruminants (cattle, sheep, goats) are the most common reservoirs: infected animals present with congenital defects, late abortions, stillbirths and endometritis, resulting in substantial economic losses [1]. Recent African data report 33% seroprevalence in abattoir workers and between 27% and 49% in pastoralists [2-4]. In Northern Tanzania between 2016 and 2017, probable acute Q fever was diagnosed in 32.5% (74/228) patients over 2 years of age, with one confirmed case [5]. In South Africa, Q fever was diagnosed in 38.3% (28/73) of adults presenting with an acute febrile illness at a rural clinic between October 2012 and June 2013 [6]. Notably, this study was reported from an area characterised by an intimate association between the human population, livestock and wildlife, through porous boundaries between communal grazing areas for cattle and goats, subsistence farming and local villages, and wildlife reserves [6].

Between 1990 and 2022, 81 outbreaks of Q fever in humans were reported globally, predominantly in high-income countries (HICs) [7]. Amongst these was an outbreak reported in Israeli tourists to Kenya, where 4/50 (8%) of travellers contracted Q fever, of whom two developed overt infection [8]. In sero-epidemiological studies, high rates of C. burnetii exposure compared with the general population were reported globally amongst veterinary staff and students [9-12]. Outbreaks have additionally been reported affecting veterinary personnel. Outbreaks in association with parturient companion animals at veterinary hospitals and animal care facilities have previously been described [13], indicative of the variety of animal reservoirs which may be implicated, and the complex associations of Q fever at the human-animal interface. In Sydney, Australia, nine veterinary personnel were confirmed as Q fever cases, following exposure to a parturient cat [13]. An incidental finding of active disease was found during a seroprevalence study of veterinary students in Spain, including 2% (5/243) who were symptomatic. In the Netherlands, 3% (4/126) of students with positive IgG titres had positive IgM I titres [14]. In southern France, an outbreak of Q fever was reported among staff and students following exposures to small ruminants at an agricultural training centre, with 18 confirmed and four possible cases reported, respectively [15].

Problem description: in this report, the authors describe an outbreak of Q fever affecting staff at a veterinary teaching hospital, following exposure to a parturient goat admitted for caesarean section in Gauteng Province, South Africa.

Sequence of events leading to the study or investigation: data for the case reports below were accessed from the treating clinician/hospital involved in care of the patient, with full knowledge and consent of each patient, between 2024 and 2025.

Methods     

Geographic setting: cases presented occurred in veterinarians employed at a veterinary academic hospital in Pretoria, South Africa.

Case definition: a case was defined as an epidemiologically linked symptomatic patient with signs and symptoms of acute fever who tested positive by serology for active Q fever.

Laboratory methods: serological diagnosis for Q fever was performed using an immunofluorescent antigen test (IFA), confirming C. burnetii serum antiphase II IgM positive, and/or elevated antiphase I IgG and antiphase II IgG IFA titres.

Environmental studies: no environmental studies were taken to examine the farm from which the goat originated, or testing of other staff or students who had potentially been exposed.

Ethics: the informed patient consent was given for these reports.

Results     

Case report 1: a 32-year-old male veterinarian, with no prior medical conditions, presented to a hospital in South Africa in December 2022, following a three-day history of progressively worsening symptoms, including persistent fever, profuse sweating, asthenia, nausea, vomiting, debilitating headaches, photophobia, and arthralgia. Approximately one month before the onset of symptoms, the patient had performed general anaesthesia on a female goat (doe) for a caesarean section (C/S) at a veterinary teaching hospital. During the procedure, he was exposed to the goat without personal protective equipment (PPE), including no respiratory protection. On admission, his physical examination revealed a fever of 39°C, marked weakness, a tachycardia (102 beats per minute), a respiratory rate of 22 breaths per minute, and blood pressure of 164/84. A nasopharyngeal PCR test for SARS-CoV-2 was negative. Arterial blood gas analysis revealed severe hypocapnia (Table 1).

Laboratory investigations revealed moderate hepatocellular injury (Table 2) and significantly raised acute-phase marker proteins, with a monocytosis (Table 3). Thoracic computed tomography (CT) revealed ground-glass opacification in the anterior basal segment of the left lower lobe (Figure 1). The patient was empirically started on acyclovir for suspected Varicella pneumonia and ceftriaxone intravenously. On day 3, the patient´s fever resolved; however, systemic malaise persisted and worsened. Given the occupational risk, additional laboratory tests were performed to exclude potential zoonotic infection, and oral doxycycline (100 mg/12h) was initiated. Extensive investigations, including blood, bone marrow, and CSF cultures and CSF PCR testing, were negative for bacterial and viral pathogens (Table 4). Notably, serology showed an elevated anti-rickettsial IgG titre (indirect fluorescent antibody [IFA]) of 1:512 (Table 5). Empirical treatment with acyclovir and ceftriaxone was discontinued. Following doxycycline initiation, the patient´s clinical condition improved. Serology and PCR results confirmed C. burnetii infection, with C. burnetii serum antiphase II IgM positive, and elevated antiphase I IgG and II IgG IFA titres (1:64 and 1:32, respectively). Doxycycline was continued for a total of 10 days. Despite the positive serological results, the patient was discharged with neither definitive diagnosis nor appropriate treatment duration. His general practitioner reviewed the laboratory findings, establishing a formal diagnosis of acute Q fever and. extended doxycycline therapy to 21 days, due to the diagnostic delay and the severity of the patient´s presentation, beyond the recommended fortnight [16]. The patient underwent outpatient follow-up with his general practitioner to monitor disease progression and to exclude chronic Q fever infection.

Case report 2: a 32-year-old male veterinarian, with no prior medical conditions, presented to a private practitioner in South Africa in January 2023, with a history of intermittent malaise and headaches. The location of the headache was consistent with sinusitis, but it radiated to the mandible. The headaches were intractable to self-administered analgesics. Approximately two weeks before the onset of symptoms, the patient had performed general anaesthesia on the same goat doe for a C/S and resuscitation of the kids. During the procedure and the kid's resuscitation, he was exposed to the doe and the kids without PPE. Nothing remarkable was found on clinical examination; however, with a confirmed diagnosis of Q fever in his colleague and other potential occupational exposures, the general practitioner ordered laboratory investigations, including Rickettsia conorii and C. burnetii IFA antibody titres. R. conorii IgG titres were positive, and C. burnetii IgG serum antiphase I and antiphase II, positive at a 1:32 dilution. These results confirmed a diagnosis of Q fever, and doxycycline therapy, 100mg, twice daily for 14 days, was initiated. The patient reported resolution of symptoms after treatment.

Epidemiological investigation: an adult Kalahari Red doe goat was presented at the production animal clinic of the veterinary hospital, where cases 1 and 2 were employed. The doe was presented with a history of non-productive parturition (dystocia), and the owner had requested a caesarean section. On presentation, the doe was bright, alert, and responsive, with a clear, mucoid vaginal discharge without any overt signs of labour. General anaesthesia was induced in the induction room, and the goat was then moved to an operating theatre where the surgery was performed. Following induction of general anaesthesia, intubation, laparotomy, and a caesarean section, twin kids, both stillborn, were delivered. Intra-operatively, the presence of an excessive volume of intrauterine fluid was observed. Post-surgery, the doe had an uneventful postoperative recovery and was discharged after 14 days. In addition to the two cases reported above, there was known exposure of various persons, within both the theatre complex and the production animal clinic, including the surgeon, nurses, veterinary students, veterinary nursing students, and orderlies. The total number of exposed persons is unknown, and to the authors´ knowledge, no one else demonstrated symptoms consistent with Q fever nor was tested for it.

Discussion     

Risk assessment for South Africa: to our knowledge, this is the first report from South Africa of an outbreak among staff at a veterinary hospital that could be linked to a doe goat treated at the hospital for dystocia. Delayed diagnosis resulted in inappropriate treatment of the first human case, including incorrect initial selection of antimicrobial therapy and an inadequate duration of treatment. This resulted in more prolonged treatment with doxycycline, compared with the current recommended treatment [16]. Further complicating management is the well-described cross-reactivity between antibodies to Q fever and other rickettsial diseases, which are common in South Africa [6,17], emphasising the critical importance of accurate and detailed history-taking in patient management. Selection of appropriate antimicrobials for treatment of Q fever and the other rickettsioses appears to be generally overlooked in Africa, possibly due to diagnostic challenges if these are not undertaken in the context of a specific study [17], emphasising the importance of capacity building for accurate diagnoses of these diseases.

Q fever was originally described in Australia among abattoir workers [1]. Vet medicine has been identified as another high-risk profession for disease acquisition. Numerous studies have confirmed the increased risk of Q fever exposure in veterinary students entering their clinical years [9,11,12,14], especially those who have direct contact with cattle and small ruminants [14]. Typically, 60% of cases are asymptomatic, but the remaining cases develop symptomatic disease. Many patients present with mild flu-like symptoms, including a cough, but a percentage will go on to develop severe atypical pneumonia with liver involvement [1]. A percentage of patients also develop a chronic Q fever, which may persist for six months to a year or longer after the acute infection [18]. Chronic Q fever may develop after mild or asymptomatic disease, or acute clinical disease that has been inadequately treated, either through undertreatment or selection of a less effective antimicrobial [19]. The most common presentation of chronic Q fever includes endovasculitis, but Q fever endocarditis is well-described. Other manifestations of chronic Q fever include osteomyelitis, meningoencephalitis, and pseudotumours in the lung and spleen [1].

The largest Q fever outbreak reported occurred in the Netherlands between 2007 and 2010, during which 35% of small ruminant farms were affected, some of which had an abortion rate of over 60%, and 4026 human cases were notified, including 749 hospitalisations and 9 deaths [20]. This outbreak was traced back to proximity to goat farms, where the goats had a history of abortions and tested positive for Q fever [20]. Stringent control measures were introduced, including culling of affected goats, controlled disposal of carcasses, pre-treatment of animal manure intended for fertilisation, and mandatory vaccination of dairy sheep and goats [20]. Cases associated with this outbreak continued to be identified years after the outbreak concluded, in patients presenting with ongoing symptoms of chronic fatigue syndrome [19]. In 2007, an outbreak was described in southern France, with 18 serologically confirmed cases identified through active case finding. Twelve of these were at an agricultural educational centre [21]. Risk factors included males and those working or living near an area where manure was being spread. A significant association was noted between eating in the centre´s canteen and positive Q fever serology (p = 0.008). The outbreak was traced to 11 of 26 flocks of sheep and goats, which tested positive for C. burnetii, specifically two sheep flocks that were shedding high levels of the organism [21]. This outbreak was associated with an abnormally hot summer, with the affected area downwind of the implicated small ruminant flocks [21].

Similarly to the outbreak described above, the French outbreak occurred during an unusually hot and dry summer period, with average temperatures exceeding 32°C in the affected areas and unseasonably low rainfall [22]. Given these associations with unusual weather patterns, cases of Q fever and other zoonoses may occur more frequently in the future, associated with a changing climate. Q fever remains an under-recognised and underdiagnosed zoonotic disease in South Africa despite mounting evidence of widespread exposure in both human and animal populations [23]. The disease is associated with potentially life-threatening and life-altering clinical outcomes; however, its true burden is likely significantly underreported due to low clinical awareness, limited access to diagnostic testing, and the absence of mandatory notification of the disease in South Africa. Q fever is excluded from the list of notifiable and controlled diseases of 2002 in South Africa [24], despite its well-reported ubiquitous and highly contagious nature, and zoonotic potential. Although seroprevalence data from South Africa are limited, available studies indicate a substantial risk of occupational and environmental exposure, particularly among veterinarians, veterinary students, abattoir workers, livestock handlers, and individuals working within the wildlife industry [10,23].

Available serological data in South Africa on humans and animals underscore the extent of exposure risk. In a study conducted on animals at red meat abattoirs in Gauteng, seroprevalence was 9.4%, 4.3% and 0.9% in cattle, sheep and pigs, respectively [25]. In a recent study in South Africa for C. burnetii prevalence, 11 tissue samples were analysed by PCR for C. burnetii from stillbirths and abortions in this study, ten of which were PCR-positive (5 caprine and 4 ovine, including a duplicate) [26]. Correspondingly, a study of abattoir workers from 16 facilities in the Free State and Northern Cape provinces demonstrated an overall IgG seroprevalence of 33%, with facility-level prevalence ranging from 8% to 62% [4]. High seroprevalence has also been reported in cattle on communal farms, including 24.3% in Limpopo province [27] and 38.4% on farms in Bushbuckridge municipality bordering the Kruger National Park, highlighting the importance of the wildlife-livestock interface in disease maintenance and transmission [28]. Human studies from the same regions reflect parallel exposure patterns, with approximately 30% of Kruger National Park game rangers and 61% of rural cattle herders in Bushbuckridge, Mpumalanga, demonstrating prior exposure to C. burnetii [6,29].

Amongst the small ruminants, it has been noted that goats have higher shedding rates of C. burnetii than sheep [30]. Reactivation of infection occurs in pregnant females and caused abortions in goats and, less frequently, sheep with high concentrations of C. burnetii found in the placentae and lochia of infected animals [31]. This has implications for outbreak control, as well as economic implications. Although culling of pregnant sheep and goats was used in the outbreak in the Netherlands [20]. Vaccination remains the most cost-effective strategy to control animal disease [32], and would be most appropriate in an African setting, where wealth is intimately associated with ownership of livestock.

This study illustrates the significant clinical consequences of a missed diagnosis. Failure to recognise Q fever resulted in inappropriate treatment and delayed initiation of doxycycline. Early initiation of appropriate antimicrobial therapy curtails disease duration and reduces the risk of long-term sequelae, whereas failure to recognise the condition may predispose patients to debilitating complications, which can be life-threatening if left untreated [16]. There is a broader issue of underdiagnosis and misdiagnosis of Q fever in Africa, driven by limited clinical awareness, non-specific clinical presentations, inadequate consideration of occupational and animal exposure during history-taking and limited access to appropriate diagnostic methods. Njeru et al. (2016) reported an increased likelihood of patients in Kenya discharged from hospital with inappropriate treatment for C. burnetii infection due to misdiagnosis, further highlighting the under-recognition of the disease in Africa [33]. Long-term fatigue was reported in an estimated 10-15% of acutely symptomatic Australian and British patients [29], underscoring the importance of early recognition and appropriate treatment. Vaccination represents a highly effective preventive strategy in humans, as demonstrated by successful Q fever vaccination programmes in Australia [29], but no programme currently exists in South Africa. Vaccine availability, cost, and the requirement for pre-vaccination screening pose significant barriers to implementation [29]. Nevertheless, given the expanding wildlife industry in Africa, increasing human-animal interactions, and demonstrated exposure across multiple sectors, vaccination of high-risk occupational groups warrants serious consideration.

Limitations: in the present case series, both cases 1 and 2 were involved in the anaesthetic management of the same production goat, which represented the only production animal case jointly managed by both veterinary anaesthesiologists, strengthening the likelihood of a common source exposure. Neither the goat, the hospital environment, nor the agricultural environment from which the goat originated was tested for C. burnetii, representing a limitation of this report, potentially weakening the causal link; however, the clinical context, a caesarean section performed on a goat that delivered two stillborn kids, is consistent with scenarios associated with increased environmental shedding of C. burnetii. The delayed diagnosis of case 1 undoubtedly had significant implications for contact tracing and testing of other cases who may have had subclinical infection, and for failure to investigate the farm from which the goat originated. Numerous veterinary personnel, including reproduction and production animal veterinary specialists, and veterinary students, were involved in the case, but the infection status of the remaining exposed personnel and their number is unknown, representing a further limitation regarding the outbreak investigation, so the attack rate is undefined. Moreover, as the cases were identified by passive methods, rather than though a procedure of active case finding, a case control study could not be conducted. The fact that two individuals underwent diagnostic testing, and both were IgG phase I and II seropositive, and case 1 was IgM phase I and PCR positive for C. burnetii, raises concern that additional unrecognised infections may have occurred. Notably, case 1, who developed severe, debilitating clinical disease, with pulmonary and hepatic involvement, reported no direct contact with foetal or reproductive tissues, underscoring the risk of airborne transmission within a veterinary hospital setting and the potential for widespread exposure among clinical personnel.

Inferences: veterinarians are at risk of acquiring Q fever. Outbreaks of Q Fever have previously been described in animal hospital staff in Europe, Australia and the United States. This is the first report from Africa.

Recommendations: due to inadequate appreciation of the importance of One Health in disease management, distinct errors were made in the diagnosis and management of both human and animal cases, emphasising the need to for enhanced One Health education and applications in South Africa going forward.

Conclusion     

These observations emphasise some critical points. Firstly, with changing disease patterns, there is an imperative to understand the implications for human infections and One Health in Africa. Health and agricultural policies in this regard may be insufficient to meet the population requirements and a realistic approach and urgent revision are needed to encourage a proactive approach to zoonotic outbreaks. This should include medical education, which should emphasise the importance of a detailed history, including animal contact, in any patient presenting with Pyrexia of Unknown Origin(PUO), as well as adequate laboratory support for the diagnosis of human and veterinary diseases. Secondly, this highlights that One Health is more than simply zoonotic infection, and the interaction between human, animal and environmental health; rather, it encompasses complex relationships that also include climate effects, the risk of exposure to different pathogens, dynamic environmental conditions, and specifically in an African context, the economic implications at both the macro- and the micro- level. Thirdly, this outbreak stresses the human role as sentinel for zoonotic diseases from an African perspective, a situation which may become increasingly frequent as the climate crisis deepens, which has implications for both human and animal health. Lastly, it serves as a critical reminder that a One Health approach, such as vaccination of animals, can prevent the development of human disease, while equally protecting animals and recognising the socioeconomic implications of controlling zoonotic disease in the African context. These considerations demand an agility in the approach to infectious diseases, whether at the human or animal interface, clearly defined policies for outbreak investigation and a proactive One Health approach, and greater awareness of these interactions going forward.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Christiaan Blignaut and Karen Helena Keddy devised the study and wrote the first draft. Christiaan Blignaut and Abdur Kadwa provided technical information. Martin Schulman contributed to the veterinary information related to the parturient goat and Theresa Blignaut was responsible for the diagnosis and management of the presenting case. All authors reviewed the manuscript, provided input into the draft and approved the final

Tables and figures     

Table 1: comprehensive list of arterial blood results sampled from case 1 upon presentation to the emergency department

Table 2: comprehensive list of serial blood results during hospitalisation and treatment in case 1

Table 3: comprehensive list of serial haematology results during hospitalisation and treatment in case 1

Table 4: cerebrospinal fluid (CSF) results during hospitalisation and treatment in case 1

Table 5: serology results during hospitalisation and treatment in case 1

Figure 1: thoracic computerised tomography of case one revealing ground-glass opacification with confluence in the anterior basal segment of the left lower lobe, small left pleural effusion, interlobular septal thickening, periportal oedema and fluid in the superior pericardial recesses

References